He did his pediatric oncology training at Dana-Farber and uh got a Master of Science in Epidemiology from the Harvard School of Public Health. Um, he's currently associate uh professor of pediatrics at Harvard Medical School and director of experimental therapeutics at Dana-Farber Boston Children's Cancer and Blood Disorder Center where he leads a program designed to bring new targeted therapies to children with cancer. Um, he also leads an active clinical and translational research program focused on patients with advanced neuroblastoma and Ewing sarcoma, um, which is what he's gonna, um, share his expertise, um, on, uh, with us today. He, um, conducts, uh, clinical trials of novel targeted agents relevant to these diseases and studies new biomarkers that improve our understanding of biology and pediatric solid tumors. Um, he serves on a number of committees including the Children's Oncology Group, the Neuroblastoma Steering Committee, the COG Bone Tumor Committee, uh, Committee, Developmental Therapeutics Executive Committee, and the American Society of Clinical Oncology Scientific Program Committee. Um, so, thank you, Doctor Boy, and look forward to your talk. Thanks so much for having me. I'm pleased to be able to share some of the work we've been doing on one of my two favorite diseases, Ewing sarcoma. And what I hope to convey, uh, is shown here, first, um, uh, making sure that we understand prognostic factors and risk groups in patients with this disease, give an overview of the, uh, general therapeutic approach that we take to patients with Ewing sarcoma. And for this audience, um, uh, understand local control approaches in patients with Ewing sarcoma. And then I am a, a drug developer, so I have to talk a little bit about the role of novel agents in the care of patients with this, with this disease. So, for this talk, no relevant industry rela relationships. There's a couple of unrelated industry relationships, and then um any agents that are discussed are investigational or off label. And in my view, that's, that's part of the problem. We, we don't have any FDA approved agents for this, uh, for this disease, so, we're working to change that. But for now, that's the, that's the lay of the land. So, I'll start first with some um background on the disease and um uh then move into some of the work we've been doing trying to improve our risk stratification in Ewing sarcoma, then move on to um uh work, uh, evaluating local control strategies in Ewing sarcoma, and then end with, um, attempts to improve our systemic therapies. So, uh, Ewing sarcoma is an entity first described way back in 1921 by a pathologist, um, uh, Doctor James Ewing, who ended up on the cover of Time magazine, which I guess in 1921 you get to do. Um, as a pathologist. Um, it's a disease that unfortunately has some unfortunate nomenclature that leads to some, some confusion. Um, so, there, um, uh, it's classic Ewing sarcoma, and then if there's evidence of neural differentiation, um, there used to be the, the moniker, um, peripheral primitive neuroectodermal tumor or peripheral PNAT. That's really with the most recent update of the World Health Organization pathology classification gone away, and we think of these now. Um, uh, we think of these now identically. Um, likewise, there's a term Askin's tumor, which is, um, uh, effectively a a chest wall Ewing sarcoma. And so, it's because of some of this terminology, um, uh, there can be some confusion about what we're, what we're talking about. And then, more recently, there's been an even more unfortunately named group of diseases called Ewing-like sarcoma that Um, uh, have very little in common biologically with, um, with Ewing's sarcoma, and I wish we had a different name for those, but I'm not going to talk about that distinct group of diseases today. Um, uh, Ewing's is typically thought about as, uh, a tumor of the bone, and that's true in 75 to 80% of cases, but in about 20%, there's, um, really no evidence of, of bone involvement, and it's purely a soft tissue, uh, soft tissue tumor. This is a classic adolescent and young adult malignancy, so you can see, uh, here the incidence per year per million as a function of age at diagnosis for boys and girls, and this is a pretty normal distribution centered right around the mid-teen years. The other interesting aspect of the epidemiology of this disease is that it's rare in, uh, people of East Asian and Sub-Saharan African ancestry, and that's true even in, um, in, um, uh, here in the United States. Um, that's a, a, a topic I won't talk about much more today, but is something under active investigation to try to understand the, the, the genetic basis for that. This is also what we consider a classic translocation associated sarcoma. So, multiple, uh, genomic landscape studies have, um, have looked at the disease and have found that it is a remarkably simple genome, characterized by, um, uh, the, the main genomic finding is a recurrent reciprocal genetic translocation. That it typically involves the EWSR1 gene and one of the um uh ETS family transcription factors, most commonly fly one. And some would argue that you can't really have Ewing's sarcoma if you don't have one of these translocations. Um, we can detect it in at least 95% of cases. Probably it's there in every, every case. And it's thought that this, um, that this fusion leads to, um, an aberrant oncoprotein. So, it, uh, incorporates the um uh transcriptional activating domain of the native EWSR1 protein, together with the DNA binding domain of fly one, and it's thought that this translocation then results in, um, uh, an aberrant transcription factor that basically is the, the founding event in the disease and, and drives all of the aberrant transcriptional program that leads to Ewing's sarcoma. You may have heard the term, um, translocation 1122. That's, um, uh, that's uh from a cytogenetic standpoint, really, um, an EWSY translocation. So, as I mentioned, this can arise either in the bone or soft tissue, and unlike some other bone tumors of, of childhood, namely osteosarcoma, um, uh, Ewing's really can arise throughout the bony skeleton, um, including, um, a fairly substantial proportion arising in the, in the bony pelvis, um, uh, long bones of the extremity, and then, as I mentioned, the, the, the chest wall. This is one of our classic small round blue cell tumors of, uh, of childhood on H&E monomorphic sheets of, uh, of these small round blue cells. The most useful immunohistochemical marker, uh, used by our pathologist is, um, CD 99. And that, um, uh, typically, uh, is, um, uh, membranous and very crisp, um, uh, staining. And if we don't see this, we actually start to get a little bit suspicious that that might, that we might be dealing with something other than a, a, a classic Ewing sarcoma. And then with the discovery of the EWS fly and related translocations, more and more molecular diagnostics are playing a role in, um, in this disease. So, um, uh, you know, previously a typical karyotype could reveal the, um, the classic 1122. Here, we favor a, um, a fish assay that's a break apart probe that, um, uh, that, uh, shows that the EWSR1 native locust has been, has been split. Some centers use PCR to detect the fusion, and then increasingly we're seeing, um, particularly here, um, uh, use of next-generation sequencing panels that identify the, um, identify the fusion. We actually looked at this in our children's oncology group open um metastatic study, just to get an understanding of, well, what, what's happening at, um, at COG centers, and we ask this because we've wondered, well, when we design clinical trials for patients with Ewing sarcoma in 2018, should those be Uh, should there be a molecular criterion to, um, uh, to enroll, or is it really a histopathologic diagnosis, which has been our, has been our trend. So, we were interested in what, what is the lay of the land. So, we started with 202 patients who had, who had enrolled, and the vast majority, as would happen here at our center, had fish on their, um, on their primary tumor. Some had, um, uh, PCR and 3 had other testing, either next generation sequencing or just a simple, um, a simple karyotype. Uh, there were 26 patients who had no testing on the primary tumor, but this was a, a metastatic population. And so there were a few patients who had testing on their, um, on their metastases. But at, at the end of the analysis, we found that there were 6% of patients who had no translocation testing. And 16% of patients where we couldn't document that they had a typical Ewing sarcoma associated, um, associated fusion, and that's a bit of a concern because some of those, some of the diseases that have variant biology, Um, uh, may not, may not respond in the same way to some of the targeted therapies that we're investigating. So, this is, I, I suspect as the field moves forward that will now start to require molecular confirmation of an EWSR1 or related fusion. So, with that background, I'll move on to how we think about risk stratifying uh new patients when, uh, when they present. And really the, the classic prognostic factors are stage, which um remains our paramount, um, prognostic factor shown in the panel on the right, where there's a, a, a, a huge difference in, in this case, relapse-free survival between patients with localized Ewing's versus newly diagnosed metastatic Ewings. Other prognostic factors that have been reported and are, are, um, generally corroborated are, um, older age appears to be an adverse prognostic factor. Um, and that's the case whether we split at age 1518, 21. Um, uh, if you're in the older category, you tend to have less favorable outcomes. Tumor sect, particularly in the pelvis, larger tumor size, and baseline LDH, which is likely just a, a, a, a marker of disease burden. So, I like Ewing's because our risk stratification, at least right now, is kind of easy. So, this is the, the Children's oncology group, um, ALL, um, uh, uh, um, risk stratification. You could be low risk, average risk, high risk, very high risk. It gets a little bit confusing, and you can have T cell disease, you're gonna have infant disease, a little bit complicated. The other disease I study is even worse, um, in, uh, neuroblastoma. This is our risk stratification table and takes into account all of these prognostic factors. But Ewing's is easy. When you're newly diagnosed, you're either metastatic or you're localized, and that's how we've set up our, um our, our clinical trials. But in some of the work I'll share with you, actually, this, this simplicity probably belies true complexity, um, and we're probably undertreating some patients with localized disease and probably overtreating some of these patients. And I won't share with you today um late effects that, that our patients face once they're um survivors of their disease, but certainly, um, if there's room to back off on therapy, um, that would be, if we have evidence to support that, that would be, uh, welcome. So, we've done a little bit of work looking at each of these uh potential prognostic factors, um, race and ethnicity, soft tissue site of the disease. We've looked at pooling some, uh, some prognostic factors into, um, prognostic subgroups, and then we've started to look, um, more and more at the tumor biology and markers of the tumor biology. So, I, um, shared with you on an earlier slide, the very low incidence, um, in, um, in black populations. And so this, um, uh, this slide shows age on the Y axis here, average annual rate per million. For, um, US black and US white populations. And you can see that this is a rare, a rare disease. So even at peak incidence, uh, age incidence, 15 to 19 years, there's, um, um, uh, uh, only 5 cases per million, and falls way off on either side of that, that peak. That's for the, the US white population, but really, um, uh, uh, 1/5 of that for the US black population. So, Um, we're working on some of the genetic epidemiology of this to try to understand what's going on, but we were also interested to know, well, how do these patients do? Is there a differential outcome by race and ethnicity? So, we utilize data from the um uh National Cancer Institute SER registry, which is a publicly available Database and just did a simple, um, uh, did a simple Kaplan-Meyer curve looking at overall survival as a function of race and ethnicity. And we saw a fairly, um, a fairly, um, uh, significant decrease in overall survival for black non-Hispanic patients compared to, uh, uh, white non-Hispanic patients. And that held, um, uh, held true in a multivariate analysis, controlling for things like stage, age, um, and primary tumor sites or some of the other prognostic factors that I already, um, uh, that I already shared. So, um, uh, we've done some additional work trying to understand is this toxicity, access to care, um, but nevertheless, an, um, an interesting observation. We were also interested in that 20 to 25% of patients who present with soft tissue Ewing sarcoma. Um, uh, for this, we leveraged, uh, uh, data available from the Children's oncology group from, uh, the last two, published randomized trials, INT 0154 and trial AEWS 0031. And for this, we looked at patients with newly diagnosed localized Ewing sarcoma, all treated with what we would consider um uh standard uh contemporary chemotherapy and um uh evaluated whether there was a difference in, um, a difference in outcome. And so, this shows the event-free survival as a, as a function of time from initial diagnosis for patients with um extraskeletal or soft tissue Ewings versus uh conventional bone Ewing sarcoma. And we did see in this um uh large data set, um, a, a, a, a difference in, uh, event-free survival that was also confirmed on a multivariate analysis controlling for other prognostic factors. But ultimately, this, um, this delta between the two groups, even though, um, uh, bordered on statistical significance for the univariate analysis and was statistically significant on the multivariate analysis. I think you would agree that there's, uh, you know, even though there's a statistical difference, there's probably not a clinically meaningful, um, meaningful difference. So, our current approach is to treat these patients with soft tissue Ewings in a similar manner to those with newly diagnosed skeletal Ewings. So, um, we were then interested to, um, to look at groups of prognostic factors. So, patients don't present with just one prognostic factor, right? They present with an age, a stage, uh, a primary tumor site, size, and so, to think about a given patient, you have to think about all of those, um, all of those prognostic factors at any, um, at any given time. And we're interested to see if we could Identify subgroups combining, um, combining each of these prognostic factors. So, we used um uh a regression technique called CART to analyze overall survival, first using data from that same SEER registry to derive risk groups, and um because we don't always know the treatment patients received as part of the SEER registry, we then applied those same risk groups to data from the Children's oncology group database where we do know the treatment that was received. And so, when you do an analysis like CART, you start off with all of your patients, and the algorithm splits on the strongest prognostic factors. So, it, it finds the, the prognostic factor that is associated with the largest um difference, in this case, in overall survival. And not surprisingly, stage is, um, uh, is the paramount prognostic factor. Um, so, metastatic and localized is the first split. After that, age, So, um, uh, above and below 18. And then, um, primary site was the next split where, um, patients with non-pelvic primaries had, um, uh, had the best outcome in patients with pelvic primaries, um, inferior. So, we ordered these then as group one with the most favorable outcome if they They had localized disease under 18 years of age and a non-pelvic primary, and then all the way down to patients with the worst outcome were those who had metastatic disease at diagnosis and were over 18 years of age. And so, um, plotted out on a Kaplan-Meiyer curve, these were the, um, uh, the, those same five groups, um, uh, from the SEER data, and you can see for patients in group one, with all of those favorable prognostic factors, we were, um, uh, at an overall survival of, uh, 75%. Um, contrast that with group five, patients with Um, older patients with newly diagnosed metastatic disease, and we were well below 25%. And then we applied these same, um, these same groupings to data from the COG and, um, uh, saw a very similar, uh, very similar breakdown. So, it's possible over time that we start to um think about our clinical trials in a little bit of a different way and start to combine some of these prognostic factors. So, for example, do we need to um take a different approach for patients with newly diagnosed localized disease with pelvic tumors who continue to have inferior outcomes. And then, um, uh, uh, turning to some of the biological markers that we've started to look at, we've been interested in potential markers of disease burden in this, um, uh, in this disease. So we've looked at micrometastatic tumor cells in the bone marrow. Detection of um of the fusion transcript um by RTPCR in the peripheral blood. And then more recently, with the explosion in cell-free DNA technology, we've been able to look not at the mRNA but at the actual fusion DNA in cell-free DNA. And so I'll share with you our findings from these, um, uh, from these studies. So I've already told you that Ewing cells are classically CD99 positive, and um they're negative for CD45, which is um uh a hematopoietic marker. So, um, actually, as part of my, um, uh, uh, my, uh, K project, I, um, decided that we should use this immuno phenotype to look by flow cytometry at bone marrow samples from patients with, um, uh, with newly diagnosed Ewing sarcoma, who didn't necessarily have clinical evidence of metastatic disease. So, to look for occult micrometastatic disease by flow cytometry. And you can see here, this is um a Ewing cell line with a a flow plot CD 99 here, CD45 here, and a very nice cluster of CD99 positive CD45 negative cells. So, we did first a proof of principle study where we spiked in known quantities of Ewing cells into um uh peripheral blood cells in this case. And you can see, um, uh, as we start to increase the amount of uh Ewing cells spiked in, our, um, population here in the CD99 positive, CD45 negative git starts to, um, uh, starts to increase. And then we started with um doing this on some patients back when I was at UCSF and showed proof of concept that indeed, in patients, we could detect this small population of uh of cells. And with those pilot data, we expanded to some of our, um, some of my friends, and then ultimately to our national Ewing circle. biology study. So, we were able to gather 109 patients with newly diagnosed Ewings, all who, um, uh, shipped us baseline bone marrow samples, and we looked to see, uh, whether there was any statistical association between the flow cytometry findings and clinical, um, uh, clinical features and clinical outcomes. So, what we found um uh in um normal bone marrow was a very, um, a very low frequency of CD99 positive, CD45 negative events, all under 0. 0.01%. In the uh Ewing's cohort, patients with newly diagnosed Ewings, we can see that the, the median in the, uh, hash marks is slightly higher, and that you've got a number of patients, uh, uh, number of patients with, um, CD99 positive, CD45 negative events, well above, um, what we saw in the normal population. And then this is actually cut off. There's a few patients, um, well, well higher. We looked to see, well, was there any association between this cell burden in the bone marrow and clinical features of the disease. So, we showed, compared with control, certainly a higher, um, a higher cell burden, but didn't see any difference based on age, primary tumor site, stage, or size of the, uh, size of the primary tumor. And then unfortunately, after years of work on this project, um, you hit the button on the, the statistical program and you get a curve like this that shows, nope, not prognostic. So, we, um, uh, so this shows the, um, event-free survival rate from time of diagnosis according to, um, our flow cytometry outcome, and you can see really no difference in, um, in, in outcome. And as we were getting ready to um uh present and publish this finding, another group, um, uh, uh led by Crystal Mackle, um, had, um, had been working on a similar approach using RTPCR to look for, um, look for circulating fusion transcript. And showed a very similar, uh, a very similar, uh, finding where the level of the fusion, um, fusion transcript does not associate with, um, risk of an event. So, then, um, uh, fast forward a few years, and the technology had evolved such that we were now able to detect, um, the, um, not the fusion transcript, but the fusion itself in circulating tumor DNA. And this is a collaboration with Brian Crompton, who has a translational research lab um at Dana-Farber, and he's developed an essay called Transeek that um uh is designed really to detect pediatric specific translocations within, um, uh, within the plasma of patients with, um, uh, with sarcomas. So, um, uh, we were able to gather, uh, a, a large group of samples from, uh, patients with newly diagnosed Ewing sarcoma, and showed first, that as proof of principle that we could detect, uh, CT DNA circulating tumor DNA in about 50% of bank samples from an unselected group of patients with newly diagnosed Ewing sarcoma. We did see higher rates of detection in patients with newly diagnosed metastatic disease, perhaps not surprisingly, and then higher rates in patients with, um, uh, pelvic tumors as well, which typically are, um, larger in size. So, we, we thought that this was, um, uh, also potentially a, a, a measure of disease burden. And then in this case, we were gratified to see that CTDNA detection is actually strongly prognostic of, um, in this case, event-free survival. So this is a Kappelmeyer curve of EFS as a function of time from initial diagnosis for patients where we could detect um CTDNA and patients where we couldn't detect CTDNA. And the remarkable thing here is that this is an, um, this is a sub-analysis, um, focused, actually, yes, this is a sub-analysis focused exclusively on patients with newly diagnosed localized disease. So, in, in my view, if this is a finding that is validated and something we're working on now, If this is a finding that's validated, this would be a very strong marker that I think we could use um uh to select patients with, with clinically newly diagnosed localized disease, but who have a predicted outcome, um, of, of, uh, 50% event-free survival. And those patients, clearly we need to, um, we need new strategies. We're also working to Um, try to understand the impact of changes in CTDNA over time on serial measurement. So, does your rate of decline in CTDNA, um, uh, correlate with clinical outcome, thinking that that might be associated with, um, chemosensitivity, for example. So, this is, um, uh, an initial finding that we're certainly intrigued by, but our, um, validation work is ongoing. And then in work that's now um uh in revision, we were interested to look at additional mutations in the tumor itself, not in the circulating tumor DNA but in the tumor itself. And I shared with you already that Ewing sarcoma has a remarkably quiet genome, but there are occasional additional mutations that are seen. And they're um not particularly, uh, not particularly common compared to um adult malignancies like melanoma, um, renal cell cancer, bladder cancer, lung cancer that have an incredibly high, um, uh, mutation count. Here, um, we have, uh, a distribution of, um, mutation load as absolute mutation count. Uh, for patients with newly diagnosed metastatic Ewings and newly diagnosed localized Ewings, and you can see, um, uh, an average of around 10 or 11, um, uh, mutations per, um, per tumor. And so, in work that Daphne Has-Kogan is um uh is leading, we then looked to see, well, is there a difference in outcome based on tumor mutational um uh mutation count. And indeed, in the entire cohort of patients with localized and metastatic disease, you can see a fairly significant split. In overall survival. And then likewise, in patients with newly diagnosed local um newly diagnosed localized disease, you can see, um, a, a split as well. So this may be a marker that we, um, study more over time and may help us to understand why some of our patients with localized disease are failing when the, the vast majority have a fairly favorable outcome. So, I'll turn now to our um uh to the treatment of the disease, and I'll start with um some of the work we've been doing looking at local control strategies in um in Ewing's sarcoma. The treatment approach, the overall treatment approach that we take to this disease is shown here. So, almost all patients start with neoadjuvant chemotherapy, and then after about 6 cycles of that, move on to the local control phase, and then complete adjuvant chemotherapy. In, uh, for local control, we have several options in, in Ewing sarcoma. So, um, uh, complete surgical resection with negative margins. Uh, definitive radiotherapy, or a combination of surgery plus, um, plus radiation. And when I see patients for Ewing sarcoma second opinions, it's almost never about the chemotherapy, cause if, if you walk into any hospital in North America, you're gonna get the same chemotherapy. But really, the, the, the question is around local control. So we, We typically will do a joint consultation with, with, um, either a, a, a, a pediatric surgeon or an orthopedic surgeon and um one of our radiation oncologists to, um, to present sort of a, um, a, a, a single opinion to the patient and family. And the, the reason for some of this, um, uh, the reason that this gets complex is that surgical resection certainly is not an option for, for all patients. So, I mentioned that there's a fairly significant proportion of viewings that will arise deep in the pelvis, and these can be, um, uh, uh, rather large tumors, and a complete, uh, complete surgical resection with negative margins would leave the patient, um, uh, with fairly significant morbidity. Likewise, there are, um, a, a subset of these that arise in the paraspinal space and um would have significant, uh, neurologic morbidity for the patient if we were to do a true oncologic resection. Um, in terms of radiation, we know that we, that this is a radiosensitive tumor, and that, um, uh, that local control is generally adequate with, um, with definitive radiotherapy. But these are, are, are teens and young adults who we hope will survive their disease for decades. And so then, um, then they're left with a significant, um, late effects, including risk of second malignancy from, um, fairly high dose radiotherapy as, um, as a young person. Um, I'm not gonna share with you the, the importance of intensive chemotherapy in this disease, but it's really critical, um, uh, critical for improving outcomes in the disease. And for that reason, we've, we've sort of backed away from combined modality, uh, local control in this disease, um, because, as you might imagine, when you're using not one but two local control, uh, approaches, there can be, um, greater chemotherapy delays. So, I, I consider local control in Ewing sarcoma to be a classic story of confounding by, by indication. So, almost every study that's been done, if not every study that's been done, um, arrives at the same Kaplan-Meyer curve. So, this is either EFS or, um, uh, overall survival as a function of time from initial diagnosis, according to mode of local control. And you'll always find that radiation Um, uh, radiation is inferior. This is an unadjusted analysis and doesn't take into account that this is not a randomized study, that patients, um, receive their local control modality for a reason. Um, uh, and so we were interested in trying to tease apart whether, um, whether this finding is real, if we control for all of the, uh, the potential confounding factors that go into the, the choice of local control. So, we leveraged data available from the Children's oncology group, um, uh, starting from the last 3 randomized, um, last 3 randomized clinical trials. And, um, uh, we're interested in patients with newly diagnosed, um, Ewing sarcoma of the bone, who had complete local control data available. And who received their local control, not at a, um, not at an outlier time, so not at, not at upfront, and not with some significant delay, but sort of in the ballpark of, of when they typically would, um, when they typically would be getting their local control. And then we understand that better chemotherapy improves local control, um, and so we wanted to exclude patients who got, um, Uh, chemotherapy that wouldn't be considered, um, standard. So, we ended up with a cohort of 465 patients with complete data, um, and relatively homogeneous with regard to timing of local control and all with viewings of the bone. And we confirmed our, our expectations that patient characteristics differ by local control group, that, that this is not a random process. So, we showed that patients who were selected for definitive radiotherapy tended to be older and were much more likely to have, um, uh, much more likely to have pelvic primary tumors. We then did an analysis, um, called the propensity score analysis that takes, um, uh, multiple, uh, potential confounding variables and distills them down into a, a score that, that gives you sort of the group average propensity or likelihood that any given patient will, um, be selected for surgical local control or be selected for radiotherapy as their, um, uh, as their, uh, local control. And it's um uh a, a convenient way when you have observational, non-randomized data to try to control for those, um, uh, try to control for those confounders. So, um, these are the results when we look at, um, event-free survival, and, um, our, um, reference group here is those patients who were selected to receive definitive radiotherapy. And in an unadjusted model, indeed, the hazard ratio for those who received radiotherapy compared to those who received surgery, um, was, um, significantly higher for, um, event-free survival. Um, and then when we controlled for, um, uh, confounders using either propensity scores or just a conventional Cox model, um, including the, the, uh, confounding variables, we saw that, um, uh, that event-free survival differences were no longer statistically significant. However, um, when we looked at local failure as our outcome of interest, we saw something quite different. So we saw on the unadjusted model, higher rates of local failure in patients who were selected for definitive radiotherapy. And when we controlled for confounders in either way, either using our propensity score approach or using um the, the variables themselves, we saw it still um uh uh more than uh double the risk of local failure. And so what can we conclude from that? Uh, I think first, um, we know that patient and tumor characteristics vary across the chosen mode of local control. This is not a random, random process, and we're not going to be able to do a randomized trial of local, local control strategies. Um, we showed that the choice of local control does not impact the risk of event-free survival, but that patients treated with definitive, um, definitive radiotherapy have a higher risk of local failure. So, based on, on this outcome, when we see a particular patient, a specific patient, We favor surgical resection whenever feasible and reserve definitive radiotherapy only for those patients where the, the morbidity of a definitive oncologic surgery is deemed too morbid. In terms of combined approach where we use surgery and radiotherapy, currently, that's just reserved for patients, um, more or less who have, um, a, a, a, a planned definitive resection and unexpectedly have a positive margin, and those patients require postoperative, um, postoperative radiotherapy. In a follow-up analysis led by Nadia Locke, a radiation oncologist at, um, uh, at the Mayo Clinic, um, we were interested to look specifically at other risk factors for local failure in, um, in the disease. And this was published in, um, uh, in the Red Journal last year. And, um, I just showed 3 panels to highlight three of the, 3 of the key, um, uh, 3 of the key factors that fell out of this analysis. So, uh, the first is the primary site, and, um, it's a little hard to read from the figure, but the, um, this group are the patients with, um, uh, pelvic primary tumors, and their local failure rate exceeds 10%. We looked as well at those patients treated with definitive radiotherapy and um Oh, I've duplicated this. All right, those are the exact same. Um, this is supposed to show, um, uh, age as well. So, age, um, over 18 years of age, um, uh, was also an adverse prognostic factor. So, we're talking now, uh, as our, um, Children's Oncology Group bone committee, um, whether we need to develop a, a, a new approach for patients with, um, uh, with particularly pelvic tumors and potentially for older patients as well. In Europe, there is more of a, um, more of a move towards combined modality local control approaches, and they have some early data to suggest that maybe in patients with pelvic tumors, they're seeing better, um, uh, better outcomes than we are here. So, um, more to follow there, but certainly a group in need of, of better therapies. So, I'll, um, I'll, I'll end with some of the work we've been doing. Uh, working on our systemic therapy. So, the workhorse of treatment of Ewing sarcoma for decades has been, um, uh, conventional chemotherapy, and all of our clinical trials to date have really worked on escalating the intensity of that therapy, either by pushing the dose, adding in new agents, or shortening the frequency between cycles of therapy to intensify, um, intensify treatment. And we have made some gains with that, but I think we're probably starting to hit a ceiling of what we're able to, um, what we're able to achieve by just continuing to push the, um, push chemotherapy. So, we've been working on some, uh, uh, systemic therapies, um, both for patients with newly diagnosed metastatic disease, a very high-risk population, and for patients with metastatic Ewings I'm sorry, relapsed Ewing sarcoma. So, um, this is a study led by my mentor Holcomb Greer, um, published in the New England Journal of Medicine in 2003. I think the only study ever to be published on this disease in the New England Journal of Medicine. And it's, um, it's because, um, it was, um, a, a, a highly successful, um, uh, randomized phase 3 trial looking at the addition of two chemotherapy agents added to standard chemotherapy for patients with, um, uh, with newly diagnosed localized disease. So the blue line are patients with newly diagnosed localized Ewing sarcoma, treated with the standard chemotherapy at the, at the time. And then the purple line is um that standard with two other agents added. And the addition of those agents, iphosphamide and etoposide, really led to an enormous improvement for patients with localized disease. But unfortunately, the patients with um metastatic disease saw no benefit of the addition of that, that chemotherapy. And other strategies for the same group have, um, have likewise not improved the outcome. So, Um, with standard therapy today, we would expect, unfortunately, a very similar curve for a newly diagnosed, um, metastatic population. So, we've become interested now in not pushing chemotherapy intensity, but adding novel biologic agents to our chemotherapy backbone. And the first such study I'm leading through the Children's Oncology Group is looking at um uh uh blocking IGF-1R as a potential therapeutic target. So, we know that IGF-1R and IGF-1 are commonly expressed in these, um, in these tumors. We also know that the fusion, um, characteristic of the disease down regulates the endogenous inhibitor of the pathway IGF-BP3, thereby increasing the net effect of overall increasing pathway activation. If you do, um, uh, mouse xenograft experiments as shown here with a drug called ganitumab, a monoclonal antibody directed against IGF-1R, you can see some, um, uh, uh, decreases in tumor volume. And this class of agents has um been studied now in the um in the clinic, so this is sort of the sentinel patient that got us quite interested about this, uh, about this strategy in Ewing sarcoma. So this is a patient who, uh, an adult patient who happened to enroll on an adult phase one study of ganitumab. And by chance had, well, not by chance, but by chance he enrolled, um, uh, but, um, had a fairly significant response. So this is his, um, CT scan at baseline showing a nodule here, and, um, a, a tumor mass here. And then after two cycles of the therapy, so 6 weeks later, um, large, largely, um, resolved lesion here and much smaller, um, uh, lesion here, and then that, that continued on over time. Uh, interestingly, his FDG PET response, um, uh, was seen much more quickly. I think this was after 2 weeks. They obtained an early PET scan and you can see, um, at baseline, significant FDG uptake in the mass, and then 2 weeks later after the first dose of genitumab, um, a, a fairly dramatic metabolic response to the agent. This led then to um a uh a large number of phase two clinical trials, looking at, um, looking with 3 different monoclonal antibodies, all directed against IGF-1R, looking at, uh, patients with relapse viewing. So these are waterfall plots showing the best change from baseline, and each bar represents a patient. So, um, a negative obviously is what we're after, so -100 would be a patient with a complete response. And, um, uh, you can see, um, uh, fairly similar pattern where you have a significant number of patients with true responses, but then some with more um more minor regressions, uh, of the disease, and then some patients who are just, um, completely resistant. Some of the trials included um a, a range of sarcoma histologies. So you can see here some patients with synovial sarcoma, rhabdomyosarcoma, desmoplastic small round cell tumor, and you can see that all of the responses more or less were patients with, with Ewing sarcoma. So, that got us even more intrigued that, that there was really a signal, um, more of a signal here. Uh, yes, by, um, what we call resist, um, resist response, or, um, uh, CT or MR response, not factoring in any metabolic response. So, given those data and the really terrible outcomes for patients with newly diagnosed metastatic Ewings, we designed COG study AEWS 1221. This is a randomized trial where patients present and enroll at um at initial diagnosis. And are randomized to our standard of care, which is um our, our standard chemotherapy with induction, local control to the primary tumor, consolidation chemotherapy, and then, because this is a metastatic population, metastatic uh radiation, or to the experimental arm where they receive that same therapy, plus ganitumab added with chemotherapy cycles, and a maintenance phase with 6 months of um uh of the, of monotherapy with ganitumab. We actually designed this study initially as a phase 2, selection design, looking with relaxed statistics for Um, a signal to suggest that we should move into the, um, newly diagnosed localized population. Um, and we did that because we didn't have, um, adequate drug supply when we started the study, and we thought that we might get maybe 2 or 3 patients a month enrolling to the trial. And we're pretty shocked to see that we're, we're enrolling closer to 5 or 6 patients per month, so, you know, Ewing's is a rare disease. Most patients don't have metastatic disease, and so this is a rare subgroup of a rare disease, but nationally, um, we've been able to find these, find these patients. So, given that rapid accrual and actually some additional drug supply that became available, we've now expanded this to a phase 3 trial with a goal of um randomizing 300 patients, and we're at like 280. So, we're in the home stretch here, and in several years, I'll be able to come back and share the results of, of this. But it's been, I think, rather telling to our, to our field, um, uh, how many of these patients are out there, if you've got a clinical trial for them, and how willing they are to participate in a clinical trial, to try to, um, uh, uh, to try to, uh, help us learn and potentially access a novel agent. And then I'd like to end with, with this slide. So, it's, um, you know, at the start of the talk, uh, I, I shared with you that we understand completely what drives this disease. It's this EWSXS or EWS fly, um, uh, fusion oncoprotein. And that, that is what we view as the linchpin of, of viewing tumors. And if we could understand More about the function of that, and get some of our medicinal chemists working on blocking the function of that, that driver, we think that we'd have a fairly active, uh, we think we'd have a fairly active agent. So, there are, um, a couple of agents that have just now entered clinical trials that are getting a little bit closer to the, what I consider the heart of the matter with this, with this disease. Um, one that, um, um, uh, one that we've just opened here, we'll, um, enroll nationally our first patient, um, uh, today on the clinical trial. We'll see how, how they do. And then there's some, um, some trials open at other centers that are trying very much to get at, um, uh, get at this, um, this driver. So, those will start in the relapse setting, and then if there's evidence of activity, we'll obviously move that up to some of our higher risk frontline, um, frontline patients. So, with that, I'll just share with you, um, uh, my acknowledgments, my mentors and collaborators, funding sources, some of the people who did the work back when I was at UCSF, the research nurses and the research coordinators, um, uh, who do a tremendous amount of work, um, on these, on these projects. And of course, the patients and families who, um, you know, have tremendous courage to enroll on these clinical trials, um, either with an experimental agent or in the case of a randomized study, just getting the, you know, randomized and, and not getting the experimental agent. So, this is the aunt of one of my patients being treated now, a patient I share with, uh, Doctor Weldon, um, and she comes in basically every week with a different t-shirt, and it's great fun. So, I'll end there and happy to answer questions. Thanks. Well, Steve, I'd like to first thank you for really a superb presentation of where we are with you and Sarcoma and the journey that we've taken so far. It's um Also, thank you for all of the, the work you're doing on the experimental agents cause when you take the Ewing's sarcoma patients that have metastatic disease, it's clearly a very dismal group. It's at least gone from 10% survival up to 20% survival, but there's still a long, long way to go. So I guess I'm not surprised that in that cohort, you're having rapid enrollment of studies because the oncologists nationwide realize. What, what the chances are and the families are very anxious to try anything new that's gonna enhance their, their agents. So is the major blocker for new development of agents industry coming up with them, industry, uh, creating the antibodies against what have been marked as potential, um, targets, or what's, what's really the rate limiting. Factor. Yeah. I'm glad you asked that. It's um, it's something we really struggle with. This is, this is not just an orphan disease, this is like an ultra-orphan disease. So, in the United States each year, I didn't share with you the, the incidence, but we probably have maybe 300 patients per year in, in the United States. So, there's effectively no market um for, for the disease, and because the biology is unique, a drug developed to target EWS fly, in all likelihood is not gonna have activity in any other disease, right? If it's really a specific inhibitor of that key driver. So, all of the things that have entered the clinic now. That I would view as sort of agents that that that target EWS fly have all come from academic labs that have been funded either by the NIH or by by sarcoma foundations or pediatric cancer foundations, and have been picked up by very small companies who are looking for orphan, you know, home run orphan drugs. Um, we don't have Big Pharma helping us with this, with this disease. Um, and so, I think we have, um, an issue with, um, with resources. I, in terms of the trial I'm leading here with, with Ewings, I didn't share with you a little bit of the, the history, but it sort of, it speaks to your question. So, all of these IGF-1R monoclonal antibodies were investigated in adults with breast cancer, lung cancer, pancreatic cancer, and Ewing sarcoma. And they, we saw a signal in Ewing sarcoma, but all of the studies in the other diseases failed. And so every company has shelved their agent. And in fact, this study almost died, and um the National Cancer Institute stepped in, and, and um Amgen had been developing uh ganitumab. Uh, uh, the NCI stepped in and said, no, you know, we need to do something for this rare disease. We are going to acquire from you your supply of ganitumab that you have said you're never going to use again, and Um, and so, they, they now have taken over the supply and um we're using it for, for the trial. So, this moved forward, not because of pharma, but because of the Children's oncology group, academia, and the NCI really pushing extremely hard um to get this, uh, to get this open. So, that's an example of, of an agent where we're sort of borrowing an agent being developed for adult cancers and trying to apply it to um uh to a pediatric cancer, and we, we struggle. Additional questions for Doctor Dubois. Jehosa. Very elegant talk. Uh, is there any association between high tumor mutation count and pelvic localization? And if not, is there any genomic signature more common in, uh, pelvic tumors? Um, we did not see, we saw higher tumor mutation count in patients with newly diagnosed metastatic disease and in patients with, um, uh, with older age, but not with pelvic, um, uh, not with pelvic tumors. Um, you know, the age, you sort of wonder, you know, is that just natural history, right? We all, over time, as we age, develop more, more mutations. So, is that, um, Is tumor mutation count really just a, a proxy for age? We did some multivariate analysis that suggests that that's not, not really the case. Um, I think the second part of your question is really, well, why do patients with pelvic tumors do worse? Is there a different intrinsic biology? So are there other, um, other things we don't understand yet about what's, what's driving those tumors? Or is it really a size phenomenon where these are, um, you know, Rather enormous tumors, uh, potentially with a very hypoxic, um, hypoxic center and maybe some of our chemotherapy, some of our, um, radiotherapy strategies that, you know, maybe don't work as well in a hypoxic environment, maybe are not, um, maybe are not as, as effective. So, um, I don't think we totally understand yet why they do worse, but it's, it's clear that that we need to change our strategy for them. Any final question for Doctor Dubois? If not, Steve, thank you so much for thanks so much.
Click "Show Transcript" to view the full transcription (47812 characters)
Comments